Automated drill bit forensics : enhancing efficiency and accuracy through image processing and machine learning

Access full-text files

Date

2023-12

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

In recent years, the automation of drilling has garnered considerable attention within both the upstream oil and gas companies and the drilling research community. Drill bit forensics, being integral to the enhancement of efficiencies and profits in the oil and gas industries, promises heightened drilling efficiency, augmented consistency, and a refined comprehension of bit damage mechanisms through automation. Nevertheless, the adoption rate of drilling automation remains sluggish, largely due to the intricate nature of drilling operations. At present, the conventional inspection and grading of bit damage by human operators is labor-intensive and susceptible to human biases. This underscores the imperative for an automated system in drill bit forensics, which would aid drilling operators and specialists in processing and analyzing bit damage data. In this dissertation, a novel systematic framework is introduced, amalgamating computer vision and machine learning techniques with domain-specific knowledge of drill bits. This framework streamlines the evaluation process from identifying various drill bit components, quantifying and categorizing cutter damage, collating positional data, to ultimately forecasting the primary causes of damage. The methodologies devised are applied to visual data of drill bits, encompassing images and videos from hundreds of different bit runs. This work delves into several innovative contributions: (1) The industry's first bit detection model that segments distinct parts of the bit; (2) A pioneering proposition to utilize video data of drill bits to expedite the automation of bit forensics; (3) A comprehensive workflow tailored for diverse bit data sources; (4) An adaptable analytical methodology for discerning the root causes of bit damage. The outcomes underscore the potential of an automated system in drill bit forensics to bolster the precision and uniformity of drill bit assessments, offering invaluable insights into drilling operations. This groundbreaking methodology lays the foundation for further advancements in the realm of automated drill bit forensics, targeting the enhancement of the overall efficacy and cost-efficiency of drilling operations.

Description

LCSH Subject Headings

Citation